It has become recognized that the clinical circumstances at the time of a first episode of VTE are the most important factor that influences the risk of recurrent VTE [1–4]. If VTE is associated with a major non-reversible risk factor, such as cancer, patients have a very high risk of recurrence if anticoagulant therapy is stopped (i.e., ≥15% in the first year). Consequently, patients with active cancer and a first episode of VTE are usually treated indefinitely . At the other end of the spectrum, if VTE is provoked by a major reversible risk factor, such as recent surgery, patients have a very low risk of recurrent VTE when anticoagulants are stopped after 3 months of treatment (i.e., ≤3% in the first year) [3,6]. If VTE was associated with a less substantial (or minor) reversible risk factor, such as a medical illness or a soft tissue leg injury, the risk of recurrence is higher than if VTE was provoked by recent surgery (i.e. ∼5% in the first year), but still low enough a risk that anticoagulants can be stopped after 3 months of treatment [3,6]. Patients who have had an ‘idiopathic’ (also referred to as ‘unprovoked’ or ‘spontaneous’) VTE have a risk of recurrence that is higher than patients who have VTE associated with a reversible risk factor, and lower than those who have VTE associated with cancer [2,3]. If the idiopathic VTE was confined to the calf veins, referred to as ‘isolated distal DVT’, the risk of recurrent VTE is about half of that for proximal DVT or PE, and three months of anticoagulation is adequate therapy for these patients [5,7,8]. Depending on how aggressively it is sought (many centers do not scan the calf to detect non-extending distal DVT), isolated distal DVT account for 0 to 25% of patients with idiopathic VTE. Therefore, most patients who are treated for idiopathic VTE have proximal DVT (i.e., involvement of the popliteal or a more proximal vein) or PE. The risk of recurrence after stopping anticoagulants appears to be the same in patients with idiopathic proximal DVT or PE, with a cumulative risk of recurrence of about 10% at 1 year, 30% at 5 years and 50% at 10 years [3,9]. As there has been uncertainty about how long to treat patients with idiopathic proximal DVT or PE, this subgroup of patients with VTE has been the focus of a number of randomized trials that have compared different time-limited durations of anticoagulation (e.g., 3 vs. 12 months), or have compared a time-limited duration of therapy with indefinite anticoagulation [8,10–14].
Summary. About half of the patients with a first idiopathic proximal deep vein thrombosis (DVT) or pulmonary embolism (PE) are expected to have a recurrent episode of venous thromboembolism (VTE) within 10 years of stopping therapy, and randomized trials have shown that this group of patients benefit from long-term anticoagulant therapy. High risks for bleeding and patient preference are compelling reasons not to treat such patients long-term. Although a number of factors are associated with a reduced risk of recurrence, they require further validation before it is appropriate to routinely stop anticoagulants after 3 months in patients with idiopathic proximal DVT or PE.
Different time-limited durations of therapy for idiopathic VTE
Studies performed over a decade ago established that shortening the duration of anticoagulation from 3 or 6 months to 4 or 6 weeks results in a doubling of the frequency of recurrent VTE during followup in most patients with VTE, and particularly those with idiopathic VTE . Whether extending the duration of treatment beyond 3 months reduces the risk of recurrence if anticoagulant therapy is then stopped has been less clear.
Four studies have compared 3 months with either 6 or 12 months of anticoagulant therapy, specifically in patients with idiopathic VTE [10,11], or in patients with either idiopathic VTE or VTE provoked by a temporary risk factor [8,15]. All four studies found that the risk of recurrent VTE was low while patients were receiving anticoagulant therapy, but increased markedly once treatment was stopped. Importantly, the risk of recurrence after stopping therapy was not lower (i.e., was the same) in patients who stopped anticoagulants at 6 or 12 months compared with those who stopped treatment at 3 months (relative risk for recurrence 0.95; 95% CI 0.72 to 1.26) . The findings of these studies, therefore, suggest that: (i) 3 months of anticoagulant therapy effectively treats the acute episode of VTE and reduces the risk of recurrent VTE to as low a value as can be achieved by a time-limited duration of therapy; and (ii) treatment beyond 3 months prevents new episodes of VTE that are not directly related to the preceding episode. The associated implications are that patients with idiopathic proximal DVT or PE should receive 3 months of anticoagulation to treat the acute event, and then a decision should be made to either stop treatment or to continue it indefinitely, with the option of subsequently stopping treatment if the patient’s risk of bleeding becomes excessive (i.e., acquires risk factors for bleeding) or if the patient decides that the burden of anticoagulant therapy becomes unacceptable . What are the benefits and risks of long-term anticoagulant therapy in patients with idiopathic proximal DVT or PE?
Indefinite vs. time-limited anticoagulant therapy for idiopathic VTE
Three trials have compared long-term anticoagulant therapy with stopping therapy in patients with idiopathic VTE who have completed at least 3 months of initial treatment [14,16,17]. Although all three studies exclusively enrolled patients with idiopathic VTE, there were differences in the study populations: the study by our group (LAFIT)  enrolled patients with a first episode of proximal DVT or PE (patients with isolated distal DVT were not eligible); the PREVENT study  enrolled patients with one (62%) or more (38%) episodes of VTE and included those with isolated distal DVT (proportion unknown); and the PROLONG study  enrolled patients with a first episode of proximal DVT or PE who had a positive D-dimer result 1 month after stopping anticoagulant therapy. Long-term anticoagulant therapy was targeted to an INR of 2.5 (range INR 2.0 to 3.0) in the LAFIT and PROLONG studies and was targeted to an INR of 1.75 (range INR 1.5 to 2.0) in PREVENT.
The findings of the three studies were consistent and impressive [14,16,17]. Allocation to long-term, standard-intensity, anticoagulation reduced recurrent VTE by about 90% (intention-to-treat analysis) and, in those who remained on anticoagulant therapy (inclusive of times when anticoagulation was ‘subtherapeutic’), the reduction in recurrence was closer to 95% [16,17]. Recurrence rates in patients who stopped anticoagulant therapy were high and consistent with the values previously described. Long-term anticoagulant therapy was associated with about a doubling of the frequency of major bleeding, with an absolute rate of major bleeding of 1 to 2% per year [14,16,17]. The overall benefit of remaining on long-term anticoagulant therapy was sufficiently convincing that, in the two studies that performed interim analyses, the studies were stopped early [14,16,17].
The findings of two other studies of long-term anticoagulant therapy for VTE reinforce the findings of these three studies. The DURAC 2 study showed that, in patients with a second episode of VTE (not exclusively idiopathic events), indefinite anticoagulant therapy reduced the risk of recurrence by over 90% compared with treating patients for just 6 months . The ELATE study, which randomized patients with one or more episodes of idiopathic VTE to long-term anticoagulant therapy with either a target INR of 1.5–1.9 or to a target INR 2.0–3.0, showed that standard-intensity therapy reduced the risk of recurrence by two-thirds compared with low-intensity therapy and that the rate of major bleeding was low and similar at about 1% per patient-year, in the two groups .
Consistent with there being a marked benefit to long-term anticoagulant therapy in patients with idiopathic proximal DVT and PE, the composite outcome of recurrent VTE, major bleeding and death, was shown to be reduced by about 50% with extended low-intensity anticoagulation compared with no extended therapy in the PREVENT study (predefined analysis) , and with standard-intensity therapy compared with low-intensity therapy in the ELATE study (‘post hoc’ analysis) . Furthermore, the combined results of randomized trials that compared extended anticoagulation with stopping therapy in patients with VTE [10,11,14,16], and the trial that compared standard-intensity therapy with low-intensity therapy , suggest that long-term therapy reduces all-cause mortality .
Translating the results of randomized trials into improved clinical practise
To realize the benefit of long-term anticoagulant therapy in clinical practise, patients with a first episode of idiopathic proximal DVT or PE who are treated long-term should: (i) be similar to the patients who were included in the previously noted trials [13,14,16]; and (ii) need to have access to good quality anticoagulant management. Therefore, not all patients with a first idiopathic proximal DVT or PE are candidates for long-term therapy. In the LAFIT and ELATE trials, about one-quarter of patients who were screened had a condition that excluded them from participating [13,16]. Levi et al.  reported that about 16% of patients who were receiving long-term anticoagulant therapy for VTE would not have been eligible for the trials that established a role for long-term anticoagulation in patients with VTE and, in the same study, showed that the risk of bleeding increased markedly with the number of exclusions that were present in individual patients.
Commitment from patients and health care providers is required to achieve good control of anticoagulant therapy. Having completed 3 months of initial anticoagulation it may be evident that, because of poor compliance or other factors that cannot be corrected, individual patients may be poor candidates for long-term therapy.
Patient preferences and the burden of anticoagulation
The burden associated with being on anticoagulant therapy differences markedly among patients. For example, Locadia et al.  found that being on anticoagulant therapy was associated with a median utility of 0.92 (where 0 is equivalent to death and 1.0 is equivalent to perfect health) in patients with a history of VTE; however, this utility was 0.77 or lower for a quarter of patients, and was 0.98 or higher for another quarter of patients (i.e., rated as better than not being on anticoagulant therapy, which had a median utility of 0.96). Consistent with these large differences in patients’ perceptions, irrespective of whether the risk of recurrent VTE was assumed to be very high or very low, 25% of patients always wanted to stop treatment and another 23% of patients always wanted to stay on treatment .
Are there patients with a first idiopathic proximal DVT or PE who should not receive long-term anticoagulant therapy?
There are two main reasons why long-term anticoagulant therapy might not be indicated in individual patients with a first idiopathic proximal DVT or PE even if their anticoagulant therapy was well controlled, and they did not mind being on treatment. First, as previously noted, the associated increase in their risk for bleeding may be high enough that it would out-weigh the benefits from reducing recurrence. Second, their individual risk of recurrence after stopping anticoagulants may be lower than for most patients with idiopathic proximal DVT or PE and, consequently, the reduction in recurrence that would be achieved by long-term anticoagulants might be too small to justify indefinite therapy.
Individualized risk of bleeding
Of factors that have been evaluated as risk factors for major bleeding during anticoagulant therapy, the following appear to have the greatest potential to be clinically useful markers of increased risk: older age, particularly after 75 years; previous gastrointestinal bleeding, particularly if not associated with a reversible cause; previous non-cardioembolic stroke; chronic renal or hepatic disease; concomitant antiplatelet therapy (to be avoided if possible); other serious acute or chronic illness; and poor control of anticoagulant therapy [22,23]. Currently, other than checking that patients would have been eligible for the trials that established a role for long-term therapy, there is no validated way to estimate and balance an individual’s increase in risk of bleeding on anticoagulant therapy with their increase in risk of recurrence if anticoagulant therapy is stopped, to identify which patients with idiopathic VTE should, or should not, be treated long-term. To add to the complexity, risk factors for bleeding may also be risk factors for recurrence. For example, in the ELATE study that compared low and standard-intensity anticoagulant therapy, patients with risk factors for bleeding at enrollment had a higher risk of recurrence during follow-up, and risk factors for bleeding appeared to increase rather than decrease the benefits of standard-intensity anticoagulant therapy . However, as the risk of bleeding increases substantially after 75 years or age, and as there is no convincing evidence that advanced age increases the risk of recurrent VTE, I am reluctant to recommend long-term therapy for patients with a first episode of idiopathic proximal DVT or PE who are older than 75 years .
Individualized risk of recurrence
As previously noted, the risk of recurrence after a first idiopathic proximal DVT or PE is estimated to be about 50% after 10 years. If it was possible to identify the 50% of patients who would not have a second episode of VTE, anticoagulants could safely be withdrawn in these patients. Current evidence suggests that female gender , low D-dimer levels one month after stopping therapy [17,25], absence of residual deep vein thrombosis on ultrasound , absence of hereditary and acquired thrombophilia , and absence of the postthrombotic syndrome  may be able to identify patients with a lower risk of recurrence (see accompanying article by Eichinger  for discussion of risk factors for recurrence in patients with idiopathic VTE). Among these factors, a low D-dimer level 1 month after withdrawal of anticoagulants currently appears to have the greatest potential as a way to identify patients with a first idiopathic proximal DVT or PE who have low enough a risk of recurrence that long-term anticoagulant therapy is not indicated [17,25]. Alternatively, as proposed by Rodger et al. , it may be preferable to combine a number of factors that are independent predictors of a low risk of recurrence (e.g., female gender and low d-dimer levels) to identify subgroups of patients with idiopathic VTE who can safely stop anticoagulants after 3 months of treatment. An additional consideration is that, as a recurrent episode of VTE is about 3-times as likely to be a PE after an initial PE compared with after an initial DVT, the risk of dying from recurrent VTE is expected to be about twice as high after an initial PE than after an initial DVT (see accompanying article by Baglin for discussion of morbidity and mortality associated with first, and recurrent, VTE. ). Although there are a number of factors that influence the risks and benefits of long-term anticoagulant therapy in patients with idiopathic proximal DVT or PE (see Table 1), I suggest that no factor (or combination of factors) is currently well enough validated as a negative predictor for recurrence that the presence of that factor would justify routinely stopping anticoagulant therapy at the end of 3 months of treatment in such patients.
|Factors that favor stopping anticoagulant therapy in patients with idiopathic proximal DVT or PE|
|History of bleeding without a reversible cause||Higher risk for bleeding|
|Chronic renal or hepatic failure||Higher risk for bleeding|
|Requirement for antiplatelet therapy||Higher risk for bleeding|
|History of non-cardioembolic stroke||Higher risk for bleeding|
|Persistently poor anticoagulant control||Higher risk for bleeding|
|Older than 75 years when VTE diagnosed||Higher risk for bleeding|
|Informed patient wants to stop||Patient preference|
|Factors that are expected to add to the benefits of indefinite anticoagulant therapy|
|Second or subsequent idiopathic event||Higher risk for recurrence|
|Positive D-dimer after withdrawing therapy||Higher risk for recurrence|
|Pulmonary embolism||Higher risk of a fatal recurrence|
|Persistent pulmonary hypertension||Higher risk of a fatal recurrence|
|Male gender||Higher risk for recurrence|
|Established postthrombotic syndrome||Recurrence may aggravate and these patient may also have a higher risk of recurrence|
|Hereditary or acquired thrombophilia||Higher risk for recurrence|
|Residual deep vein thrombosis||Higher risk for recurrence|
Dr. Kearon is supported by the Heart and Stroke Foundation of Ontario.
Disclosure of Conflict of Interests
The authors states that he has no conflict of interest.